To provide a content data regeneration device capable of outputting content data rich in variation in response to a mental image of an occupant, without requiring a complicated operation by the occupant.

An information collecting part 12 collects information on a date and a time, a weather condition, a place, occupant constitution, a destination, a riding purpose, an operation situation, a traffic situation, a bodily physical condition, a regeneration frequency, a preference, a profile and the like, and a mental image estimation part 14 estimates the mental image of the occupant, based on the information collected therein. A menu setting part 13 sets a content of an imaginary button prepared on a touch panel 32 based on the estimated mental image to collect further information required for estimating the mental image of a user by a simple user input. A data selection part 15a of a regeneration processing part 15 selects the content data suitable for the estimated mental image to prepare a play list.